Microsoft has officially launched its Windows ML platform for general production use, facilitating the integration of artificial intelligence into Windows applications. According to a blog post from Microsoft, the platform is compatible with all devices operating on Windows 11 24H2 or later and offers a framework for local AI inference, designed to create more “responsive, private, and cost-effective” AI applications.
Microsoft elaborated on the workings of Windows ML, explaining that hardware partners will develop and manage execution providers. Windows ML will oversee the distribution, management, and registration of these providers to efficiently perform AI tasks on devices, which acts as a hardware abstraction layer for developers. This ensures that applications can leverage the most suitable hardware for specific tasks, utilizing GPUs for demanding operations, NPUs for energy efficiency, and CPUs for versatility.
Initially unveiled in 2018 for Windows 10, Windows ML has undergone extensive development. Following its testing phase, various software developers such as Adobe, McAfee, and Topaz Labs are now in the process of integrating Windows ML into their upcoming products.
Adobe’s upcoming enhancements for its Premiere Pro and After Effects software will utilize local NPUs to support features like semantic search, audio tagging, and scene edit detection. Meanwhile, McAfee plans to employ Windows ML to identify deepfake content and scams on social networks. Topaz Labs has already implemented this platform to create AI-driven image editing capabilities in Topaz Photo. With its focus on simplifying AI development for Windows applications, Microsoft aims to inspire more innovations that could attract users to its operating system.
